Water quality failures in distribution networks ...€¦ · Fire 1 in 82,977 (0.0000012) Bioterrorism 1 in 56,424,800 (0.00000002) Food poisoning 1 in 56,424 (0.00002) Understanding

Water quality failures in distribution networks - understanding risk

Water quality failures in distribution Water quality failures in distribution networks networks -- understanding riskunderstanding risk

Rehan SadiqRehan Sadiq

Institute for Research in ConstructionInstitute for Research in Construction

National Research Council CanadaNational Research Council CanadaOttawa, OntarioOttawa, Ontario

Gander, Newfoundland September 2004

Outline

•• Water quality in distribution networksWater quality in distribution networksDeterioration mechanismsDeterioration mechanisms

Water quality monitoringWater quality monitoring

Water quality managementWater quality management

•• Risk analysisRisk analysisUnderstanding riskUnderstanding risk

Risk assessment & managementRisk assessment & management

•• An exampleAn example

What is Water Quality?

“collection of upper and lower limits on “collection of upper and lower limits on selected water quality indicators” selected water quality indicators”

PhysicalPhysical: : Temperature, turbidityTemperature, turbidity

ChemicalChemical: : Nutrients, metal, pH, organicsNutrients, metal, pH, organics

AestheticAesthetic: : Color, taste, odorColor, taste, odor

MicrobiologicalMicrobiological: : Bacteria, viruses, protozoa Bacteria, viruses, protozoa

Drinking Water Laws, Regulations and Guidelines

Protects public health

“... concentrations do not result in any significant risk to human health, over a lifetime consumption”

… required to perform regular monitoring to meet regulatory limits

Commonly monitored indicators of water quality

Common parameter Purpose

Alkalinity Corrosion controlDissolved oxygen Corrosion control; Detection of dead-end mainsFluoride Water quality monitoringNitrate Water quality monitoringpH Water quality monitoringPhosphate Corrosion control; Water quality monitoringResidual disinfectant Water quality monitoringSpecific organic Water quality monitoringTemperature Water quality monitoring; Flow managementTurbidity Water quality monitoring

Water Quality Failure

“an “an exceedenceexceedence of one or more water of one or more water quality indicators from quality indicators from specific specific regulationsregulations, or in the absence of , or in the absence of regulations, exceedence of regulations, exceedence of guidelinesguidelinesor or selfself--imposed limitsimposed limits driven by driven by customers’ needs”customers’ needs”

Sources of Water Quality Failure

Permeation

Permeation

IntrusionIntrusion

Leaching (and corrosion)Leaching (and corrosion)

Biofilm, Biofilm, regrowthregrowth

Water treatment Water treatment deficiencydeficiency

Physico-chemical Biological Aesthetics

Effects on waterquality

Back flow

Cross-connection

In flow

High breakagefrequency rate

Internal corrosionand leaching

Broken pipes anddeteriorated gaskets

Maintenanceevents

De-pressurizedand low-

pressure pipes

Contaminatedsoil

Contaminatedgroundwater

Tuberculation Formation ofbiofilm

Water qualityfrom

treatmentplant

Injuredbacteria

AOC, TOC,Nutrients

Residualdisinfectant

DBPs In-linedisinfection

pH, alkalinity,DO, etc.

Detachmentand sloughing

Regrowth andprotection of

microbes

Transients

Water quality failures in NL*

Indicator Observed StandardColor (TCU) 2 – 165 (36) 15

Turbidity (NTU) 0.06 – 4.52 1

Lead (mg/L) 0.001 – 0.101 0.100

THM (µg/L) > 20% water supplies(≈80K population)

100

Giardia 8 cases

5% boiled water advisories due to microbiological water qualityfailures (*Source to tap – water supplies in Newfoundland & Labrador, 2001)

Public UtilitiesSafety SafetyFree of excess chlorine residual

Maintaining chlorine residual

Taste and odour Taste and odourGood appearance Corrosion controlUniform water quality DBP formation

Rank of Major Water Quality Issues

Perceived riskPerceived risk Actual riskActual risk

Water Quality Monitoring - benefits

•• ReducesReduces public health public health risk by early detectionrisk by early detection

•• MeetsMeets legislated legislated requirementsrequirements

•• HelpsHelps to take decisions to take decisions for O & M activitiesfor O & M activities

Water Quality Monitoring - benefits

•• IncreasesIncreases consumer confidence consumer confidence (reduces perceived risk) (reduces perceived risk)

•• DevelopsDevelops water quality baseline datawater quality baseline data•• ProvidesProvides a proa pro--active approach to deal active approach to deal

with emerging water quality issueswith emerging water quality issues

Bacteriological Monitoring in NL

No distribution system or very smallsystem serving less than 100 people

1 sample/month

Distribution systems servingpopulation < 5,000

4 samples/month

Distribution systems servingpopulation 5,000 to 90,000

1 sample/1,000/month

Distribution systems servingpopulation > 90,000

For 90,000 plus

one sample/additional10,000/month

Disinfectant Monitoring in NL

... after a minimum 20 minute contact time, shall contain a residual disinfectantconcentration of free chlorine of at least 0.3 mg/L, or equivalent CT value.

Detectable free chlorine residual must be maintained in all areas in the distribution system.

Water Quality Monitoring - implementation

•• Decision on water quality indicators, Decision on water quality indicators, monitoring locations, frequency and monitoring locations, frequency and sampling techniquessampling techniques

•• Management and Management and reporting of collected reporting of collected datadata

Water Quality Monitoring - implementation

•• Incorporation of eventIncorporation of event--driven monitoring driven monitoring in the programin the program

•• Establishment of partnerships with the Establishment of partnerships with the community to monitor water qualitycommunity to monitor water quality

•• Development of response protocols for Development of response protocols for monitored data and maintenance, and monitored data and maintenance, and procedures to update programprocedures to update program

Water Quality - management

•• Multiple barrier approach Multiple barrier approach -- defensivedefensive

•• Total water quality management Total water quality management -- TWQMTWQM

Source water protection

Traditional water treatment

Disinfection O & M of water distribution

Monitoring Public awareness

Physical barriers Virtual barrier

Water Quality - management

•• HACCP HACCP -- (Hazard Analysis Critical Control Points)(Hazard Analysis Critical Control Points)Hazard analysisHazard analysis

Critical control pointsCritical control points

Critical limitsCritical limits

Monitoring Monitoring

RiskRisk--based corrective actionsbased corrective actions

Verification & validationVerification & validation

Record keeping & documentationRecord keeping & documentation

Water Quality Management Framework

Waterquality

indicator

Water qualitydeteriorationmechanisms

MitigativeactionsPipe

deteriorationmechanisms

Site specific factors

Water quality management

Operationand design

Pipecharacteristics

Risk

refers torefers to“joint probabilities of an “joint probabilities of an occurrenceoccurrence of an event of an event

and its and its consequencesconsequences” ” ((Lowrance, 1976)

“a triplet of “a triplet of causal scenariocausal scenario, , likelihoodlikelihood, and , and consequenceconsequence”” (Kaplan, 1997)

Risk AnalysisRisk AssessmentRisk Assessment

WhatWhat can go wrong?can go wrong?What What is the is the likelihoodlikelihood that it will go wrong?that it will go wrong?WhatWhat are the are the consequencesconsequences??Risk ManagementRisk ManagementWhatWhat can be done?can be done?WhatWhat options are available and options are available and whatwhat are the are the

associated tradeoffs in terms of cost, risks, and associated tradeoffs in terms of cost, risks, and benefits?benefits?

WhatWhat are the impacts of current management are the impacts of current management decisions on future options?decisions on future options?

Source §Annual risk of mortality

Heart disease 1 in 397 (0.0025)**Cancer 1 in 511 (0.002)Accidents 1 in 3,014 (0.0003)Alcohol 1 in 6,210 (0.00016)Suicide 1 in 12,091 (0.000083)Homicide 1 in 15,440 (0.000065)Fire 1 in 82,977 (0.0000012)Bioterrorism 1 in 56,424,800 (0.00000002)Food poisoning 1 in 56,424 (0.00002)

Understanding Risk

Cause of death is cancer for every 511th death in any year

Human Health Risk Assessment

•• Hazard identificationHazard identification•• Exposure assessmentExposure assessment•• Toxicity assessment (doseToxicity assessment (dose--response)response)•• Risk characterizationRisk characterization•• Risk communicationRisk communication

Hazard Identification

•• Examines data on contaminants Examines data on contaminants detected during monitoring and detected during monitoring and emphasizes those of concernemphasizes those of concern

•• Requires knowledge of source of Requires knowledge of source of contamination, concentration of contamination, concentration of contaminants and transport contaminants and transport mechanisms, i.e., how they reach mechanisms, i.e., how they reach the receptorthe receptor

Exposure Assessment

•• Source and release mechanismsSource and release mechanisms•• Transport, transfer and Transport, transfer and

transformation mechanismstransformation mechanisms•• Exposure point Exposure point •• ReceptorReceptor•• Exposure routeExposure route•• Estimation of chronic daily Estimation of chronic daily

intake (CDI)intake (CDI)

Toxicity Assessment

Cancer riskCancer risk: : Determine slope factor, SF Determine slope factor, SF (mg/kg(mg/kg--day)day)--11 , , from dosefrom dose--response curve response curve

NonNon--cancer riskcancer risk: : Determine reference dose,Determine reference dose, RfD RfD ((mg/kgmg/kg--dayday)) from dosefrom dose--response curveresponse curve

0 40 80 Dose

100%

50%

0

Res

pons

e

Acceptable level

Reference dose

Risk Characterization

ToxicityAssessment

Risk Characterization

Exposure Assessment

Hazard Identification

Risk Characterization

•• Unit Unit cancer cancer riskrisk == CDICDI ×× SFSF•• Number of a cancer cases (over life span)Number of a cancer cases (over life span)

== Unit Unit cancer cancer risk risk ×× populationpopulation•• Hazard index,Hazard index, HIHI == CDICDI // RfDRfD•• NonNon--cancer riskcancer risk == pp((HIHI > > 1)1)

Uncertainty

Risk estimates are highly uncertain due to:Risk estimates are highly uncertain due to:Extrapolation of doseExtrapolation of dose--response curve response curve NonNon--availability of data on exposure availability of data on exposure assessmentassessment

Uncertainty

Type IType IVariability,Variability, stochasticitystochasticity

related to natural related to natural heterogeneityheterogeneity

Type of uncertainties

Type IIType IIEpistemic, systematic, Epistemic, systematic,

uncertainty related to partial uncertainty related to partial ignorance, subjectivity and ignorance, subjectivity and

vaguenessvagueness

Uncertainty - based Methods

•• Probabilistic risk analysisProbabilistic risk analysisMonte Carlo simulations (higher order MCS)Monte Carlo simulations (higher order MCS)First order reliability methodsFirst order reliability methods

•• FuzzyFuzzy--based methodbased methodFuzzy arithmetic and possibility theoryFuzzy arithmetic and possibility theory

•• Interval analysisInterval analysis

CommunicatedCommunicated

Risk communication

Actual Actual (estimated)(estimated)

PerceivedPerceived

Risk gapRisk gap PerceivedPerceivedRisk gapRisk gap

Actual Actual (estimated)(estimated)

Non communicatedNon communicated

Ris

k Sc

ale

Ris

k Sc

ale

Example - DBPsChlorine residual + Chlorine residual +

NOM (NOM (natural organic matter) natural organic matter)

THMs (THMs (Chloroform Chloroform (CHCl(CHCl33), ), ……))

CCClCl

ClClClCl

HH

“Disinfection ranks with the discoveryof antibiotics as one of the majorpublic health accomplishments

of the 20th century”

Risk-risk tradeoffR

isk

Disinfection level

DBPsMicrobial

Regulatory regime

“In terms of risk, chlorination has allowed people to live long enough to worry about cancer”

Example - DBPsTHM = f (NOM, pH, Br, Dose, Temperature, time)

Dose ↑ THM ↑ (Treatment related)NOM ↑ THM ↑ (Source WQ & treatment related)Temp. ↑ THM ↑ (Environmental)Time ↑ THM ↑ (Design, O&M related)pH ↑ THM ↑ (source WQ & treatment related)

Health risks of DBPs are uncertain, they should be reduced whenever economically feasible

Strategies to reduce DBPs

• Control/surveillance to reduce both DBP precursors and microbial contaminants in source water;

• optimize all treatment processes, to ensure that concentrations of disinfectant are adequate;

• use of alternative disinfectants; and• reduce water age in distribution system.

Health Issues related to THMs

•• Linked to small increase in risk ofLinked to small increase in risk of bladder bladder cancer cancer andand colorectal cancercolorectal cancer

•• Some investigations link to heart, lung, Some investigations link to heart, lung, kidney, liver, andkidney, liver, and central nervous system central nervous system damagedamage

•• Other studies link THM toOther studies link THM to reproductive reproductive problems, problems, including miscarriagesincluding miscarriages

Categorization of contaminants for human health risk assessment

A: Human carcinogen; B: Probable human carcinogen (B1 and B2 represents two levels of B); C: Possible human carcinogen; D: Not Classified; E: Evidence that its is non-carcinogenic

Human evidence

Sufficient Limited InadequateNo

data

Evidence of no effect

Sufficient A A A A ALimited

B1 B1 B1 B1 B1Inadequate B2 C D D DNo data B2 C D D EEvidence of no effect B2 C D D E

Animal evidence

Comparison of Total THM (ppb)

Parameter Nitrate (mg/L)

THM (ppb)

§ WHO (1993) 45

US EPA (2001) 45 80Health Canada (2001) 45 100Australia - New Zealand (2000)

50 25098/83/EC 100UK (2000) 50 100§ ratio of individual THMs to guideline values should be less than 1

14

1≤∑

=i i

i

WHOTHM

Health Issues of THMs

Class of DBPs Compound Rating Detrimental effects

Trihalomethanes (THM) Chloroform B2 Cancer, liver, kidney, and reproductive effects

Dibromochloromethane C Nervous system, liver, kidney, and reproductive effects

Bromodichloromethane B2 Cancer, liver, kidney, and reproductive effects

Bromoform B2 Cancer, nervous system, liver and kidney effects

Defining input parameters for cancer risk model

DeterministicDeterministic Monte CarloMonte Carlo

FuzzyFuzzy--basedbased IntervalInterval

CDI = (IR × C × ED × EF) / (BW × AT)Unit cancer risk = CDI × SF

IR = intake rate (L/Day)C = chloroform concentration (mg/L)BW = body weight (kg)AT = averaging time (days)EF = exposure frequency (day/yr.)ED = exposure duration (yr.)SF = slope factor (mg/kg-day)-1

CDI = chronic daily intake (mg/kg-day)

Input parameters for cancer risk modelInput parameters for cancer risk model

How is risk determined?

Cancer risk cases in a million (over life span)

Statistical parameters

Maximum likely Minimum Maximum

Deterministic 10 - -MCS 7.4 0.34 69Fuzzy (2 - 45) 0.00087 660Interval - 0.00087 660

0

50

100

150

200

0

0.2

0.4

0.6

0.8

1

10-7 10-6 10-5 10-4

Freq

uenc

yC

umulative probability

Risk

Probability distribution (PDF)

Cumulative distribution (CDF)

0.34 × 10-6 7.4 × 10-6 69 × 10-6

Monte Carlo Simulations

0

0.2

0.4

0.6

0.8

1

1.2

10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3 10-2

Poss

ibili

ty

Risk

DeterministicInterval analysis

Monte Carlo

Fuzzy-based

Comparison of different methods

Summary

•• Water quality can be deteriorated through Water quality can be deteriorated through various sources in distribution networksvarious sources in distribution networks

•• The risk analysis for water quality failures in The risk analysis for water quality failures in distribution networks is a complex process anddistribution networks is a complex process and

•• Risk estimates are highly uncertainRisk estimates are highly uncertain

Publications1. Sadiq, R., Rajani, B., Kleiner, Y. (2004). Risk analysis for water quality deterioration in

distribution networks, Evaluation and Control of Water Loss in Urban Water Networks, Valencia, Spain, http://irc.nrc-cnrc.gc.ca/fulltext/nrcc47067/nrcc47067.pdf

2. Sadiq, R., Kleiner, Y., Rajani, B. (2004). Aggregative risk analysis for water quality failure in distribution networks, Journal of Water Supply Research and Technology : Aqua, 53(4): 241-261. http://irc.nrc-cnrc.gc.ca/fulltext/nrcc46269/nrcc46269.pdf

3. Sadiq, R., Rodriguez, M.J. (2004). Disinfection by-products (DBPs) in drinking water and the predictive models for their occurrence: a review, Science of the Total Environment, 321(1-3): 21-46. http://irc.nrc-cnrc.gc.ca/fulltext/nrcc44499/nrcc44499.pdf

4. Sadiq, R., Kleiner, Y., Rajani, B. (2003). Forensics of water quality failure in distribution systems - a conceptual framework, Journal of Indian Water Works Association, 35(4): 1-23.http://irc.nrc-cnrc.gc.ca/fulltext/nrcc46742/nrcc46742.pdf

5. Sadiq, R., Rodriguez, M.J. (2004). Fuzzy synthetic evaluation of disinfection by-products - a risk-based indexing system, Journal of Environmental Management, 73(1):1-13. http://irc.nrc-cnrc.gc.ca/fulltext/nrcc46632/nrcc46632.pdf

6. Sadiq, R., Rajani, B., Kleiner, Y. (2004). Probabilistic risk analysis of corrosion associated failures in cast iron water mains, Reliability Engineering and System Safety, 86(1): 1-10.

http://irc.nrc-cnrc.gc.ca/fulltext/nrcc45730/nrcc45730.pdf

http://www.nrc.ca/irc/uir/bu/index.html

URBAN INFRASTRUCTURE PROGRAMBuried Utilities Research

Visit us at ...